Seismic Detection of Buried Explosive Hazards
When seismic energy is induced in the ground, waves propagate in all directions.
These waves interact with buried explosive hazards in particular ways that can be measured on the surface using non-contact vibrometry. CoVar is pioneering innovative methods for analyzing this vibrometry data, enabling imaging and detection of buried explosive hazards with high probability of detection and low false alarm rates. These types of advanced sensor systems for explosive detection are keeping the warfighter safe and improving our force's maneuverability.
All types of buried explosive hazards (landmines, weapons caches, improvised explosive devices) represent some of the US Army’s foremost tactical and operational challenges. CoVar is developing a number of technologies to improve detection of these threats.
CoVar is contracted with Communications-Electronics Research, Development and Engineering Center (CERDEC) Night Vision Electronic Sensors Directorate (NVESD) to develop signal processing and visualization algorithms for the detection of buried explosive hazards using seismic non-contact vibrometry data. This project is focused on detection using standoff optical vibrometry of the earth's surface to infer the presence of threats by analyzing the interaction of active seismic waves and underground structure. Our seismic processing work makes use of novel sensor platforms to measure, analyze, and exploit these interactions.
CoVar has developed physical and statistical models of the subsurface phenomena to better understand how buried explosive hazards can be inferred from surface vibration data. These models drive data collection recommendations and, ultimately, sophisticated feature extraction and classification techniques for processing this seismic-optical data. Once processed, autonomous detection algorithms flag the buried hazards and the data can be visualized for a better understanding of what is buried in the ground.
CoVar is advancing the latest methods for mixed modal sensing, such as temporal-spatial processing, optical flow as applied to mechanical wave propagation, and spectral analysis. By leveraging our physical understanding and modeling expertise to explore the underlying phenomenology, we can better develop and design signal and image processing techniques which ultimately generate optimal detection performance.
Distribution Statement A: Approved for public release